> On July 14, 2017, 10:16 p.m., Benjamin Mahler wrote: > > src/slave/slave.hpp > > Lines 796-797 (original), 798-799 (patched) > > <https://reviews.apache.org/r/60867/diff/1/?file=1776769#file1776769line798> > > > > IIUC, this was supposed to represent the resources allocated the > > executor and is now obviated by your function, so it should be removed. > > > > We don't have a clear definition of "consumed" here, my understanding > > is that it meant the resources "allocated" to the executor.
Fixed in a separate [patch](https://reviews.apache.org/r/60907/). - Andrei ----------------------------------------------------------- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/60867/#review180593 ----------------------------------------------------------- On July 17, 2017, 12:09 p.m., Andrei Budnik wrote: > > ----------------------------------------------------------- > This is an automatically generated e-mail. To reply, visit: > https://reviews.apache.org/r/60867/ > ----------------------------------------------------------- > > (Updated July 17, 2017, 12:09 p.m.) > > > Review request for mesos, Benjamin Mahler and haosdent huang. > > > Bugs: MESOS-6441 > https://issues.apache.org/jira/browse/MESOS-6441 > > > Repository: mesos > > > Description > ------- > > This allows us to call a method to get the executor's allocated > resources, instead of call-sites having to each iterate over the > executor's pending and launched tasks. It turns out that most > call-sites forgot to account for the pending tasks, and accounting > for the pending tasks is non-trivial! > > > Diffs > ----- > > src/slave/slave.hpp 8bb03ecd86bfd87dfd27a800910130aec04e0919 > src/slave/slave.cpp adbe65fbb7c555b098eaae228c7277402452d8a2 > > > Diff: https://reviews.apache.org/r/60867/diff/2/ > > > Testing > ------- > > make check > > > Thanks, > > Andrei Budnik > >
